Coin Sorting Machine

Step 1: Brainstorm solutions

Step 2: ideate

I looked for inspiration from the real-world:

This lazy susan (to the right), gave me ideas for a mechanism I could use in my coin sorter. I realized that I could sort my coins by pushing them along a circular turntable, like these two inside the cabinet, that rotates with the movement of some lever and snaps back into place with the help of a spring. The turntable on the coin sorter would need different sized slots to allow different coin types to fall through each one into different exit chutes.

Then I chose three of my concepts, drew diagrams and ranked them based on the following 2 criteria:

Likeliness to function well
Simplicity of design

Based on my rankings (see table on the right), I chose to move forward with design #1.

I drew a more detailed diagram of my design, breaking it into 3 subsystems:

The entry chute
The rotating plates
The exit chutes

Step 4: prototyping & Testing phase II

(3 subsystems)

I created a higher-fidelity prototype next, integrating these three subsystems:

The rotating plates (continued from my initital prototype)
The funnel
The spring

From below

Subsystem 1: ROTATING PLATES

Driver Question 1: What hardware will allow for the plates to rest upon one another with the bottom one stationary and the top one rotatable?

Metric for success 1: Does the top plate rotate easily around the stationary bottom plate?

I decided to use the pivot assembly (see below), however, the only flanged bushings available in room 36 were too tall to fit two through the rotational link (the rotating plate), which had to be the thickness of acrylic. To solve this problem, I laser cut a washer out of acrylic to put on top of the hole and stuck one of the flanged bushings through that, effectively lengthening the hole.

Top plate

Heat insert:

Test results:

It worked well! I added a nut below both plates to hold the shoulder bolt in place, and my makeshift washer worked perfectly. My top plate rotates easily around.

Both holes fit perfectly on the first try. The flanged bushing was a nice tight fit in the hole and the heat insert was easy to melt in (that process seems to allow room for error).

Step 3: prototyping & Testing phase I

(1 subsystem)

I chose 1 subsystem to focus on first - the rotating plates - and built a rapid prototype of this subsystem with low-fi materials (cardboard, paper, and a pushpin).

I tested my subsystem prototype using the following driver question and metric for success:

Driver Question: What slot size range will allow a penny to fall through, but not a nickel or a quarter?
Metric for success: Does the penny fall through the hole as it passes over, but not the nickel or quarter?

Step 5: prototyping & Testing phase III

(ALL subsystems)

Subsystem 1: ROTATING PLATES (continued)

I realized after making my higher-fidelity prototype of this subsystem that I was supposed to have hardstops for the spring (so it can only rotate from a certain point to another point). So…

Driver question 3: What adjustment to my initial prototype will limit the rotational motion of the top plate to make hardstops for the spring (so it only moves as far as it needs to)?

Metric for Success 3: Is the rotational distance constant every time the mechanism is activated? Is the plate prevented from rotating further than the correct spring distance?

At first, I just edited my original prototype by slicing a piece out of the top plate and adding a shoulder bolt sticking out of the bottom plate to block its motion, as pictured below. That worked well, but then I realized that we were required to convert translational motion to rotational motion, so I had to use gears, and my whole subsystem design changed.

3D printing my tubes

In my last funnel prototype, quarters were getting stuck. So…

I designed and 3D printed three template holes to determine the correct size for the exit hole of my funnel.

Driver question 2: What hole size range for the base of the funnel will allow a quarter to fall through without getting stuck?

Metrics for Success 2: Do all coin sizes (up to quarter) fall through the hole easily without getting stuck?

In the last funnel prototype, quarters were getting stuck:

I realized that my fundamental design idea of screwing the funnel into the plate and moving it over all the holes DOESN’T WORK, because all the coins would come out at once (nothing would prevent more than one coin from releasing during each actuation) and they’d fall in the wrong holes if a smaller coin exited when it was rotated past that coin’s correct hole.

So I rethought my design, trying to figure out how to attach my funnel to the base plate, so that the rotating plate would come under it, grab one coin, and then move it over the four holes. I also reprinted my rotating plate with the thinner (.118”) acrylic, so that only one coin could be released at a time (it’s thinner than a dime).

I designed a bracket to screw into the tabs at the base of my funnel and attach the funnel to the base plate —>

Five concept sketches

Initial brainstorm

New concept! I scrapped that idea and decided to do a simpler version, with bins to collect the coins that drop from the holes, rather than tubes, and only three walls so I could access the collecting bins from inside the box:

New concept! (above)

Refined biomechanics diagram

Test Results:

I made a little turntable with a hole that rotates around another plate with a hole (which in later designs will have 4 holes - 1 for each coin type). My prototype was successful - my turntable plate drags a coin around until it falls into the hole. I used a penny to measure my hole, so it’s an exact fit: 1.9cm. The penny falls through as it passes over the hole when you rotate the top plate, but nickels and quarters do not: success! I’ll have to put the larger holes behind the smaller holes to avoid smaller coins falling through prematurely.

New funnel design attached to rotating plate

Bracket to attach funnel to base plate

Test results:

I only 3D printed one funnel, and it worked well! However, since I made the mouth of the funnel the exact width of a quarter, quarters get stuck in it (it’s a friction fit! - see photo above). For my next iteration, I’ll test to find the ideal sized exit hole.

Subsystem 3: SPRING

Driver question 2: What range of hole size will friction fit the flanged bushings through the rotating plate? What range of hole size will fit the heat insert threads through the stationary bottom plate?

Metric for success 2: Does the flanged bushing fit tightly in the top plate center hole? Does the heat insert fit in the bottom plate center hole?

I used the McMaster product details to find the appropriate hole sizes for the flanged bushing and the heat insert.

From above

Subsystem 2: FUNNEL

Finally, I began working on my high-fidelity working prototype, separating it into 4 subsytems:

The rotating plates (continued)
Box and Exit chutes
The funnel (continued)
The spring (continued)

Driver question: What funnel shape & size range will allow the funnel to hold 15 coins and allow the coins to exit the funnel one by one in a neat stack?

Metric for success: Does the funnel hold 15 coins? Do the coins exit one by one in a neat stack?

I decided to make the funnel 2x2 inches, so it’s more than big enough to hold 15 coins. I made the mouth of the funnel the size of a quarter so any coin can fit through, and the thickness of a quarter so only one fits through at a time.

Bottom plate

final DESIGN:

Driver question: What hardware will allow for the spring to attach one end to the rotating plate and the other to a stationary part? What range of spring extension/contraction will allow for the top plate to rotate the appropriate distance to cover all four differently sized holes in the bottom plate, and then spring back into place to reset?

Metric for success: Does the spring attach one end to the rotating plate and the other to a stationary part? Can the rotating plate be moved over all four holes, and does the spring reset it back into place?

I decided to use a heat insert and a nut to hold the screw (that the spring can loop over) in place from both sides on the rotating plate. I used the McMaster product details to find the appropriate hole sizes for the heat insert. I used a double nut around the screw that holds the other end of the spring to a stationary object (for this iteration I just used a cardboard base to attach everything to).

Spring in action:

Previous rotating plates prototype

First, I plan to find a way to attach the funnel to the rotating plates:

Friction fit with the coin hole in the top plate so the funnel rotates with the top plate, or
Attach it to the stationary base of my system somehow and have it hover over the rotating plate without rotating with the plate

I also plan to create a base for my system, like a clear box or cylinder, to attach to the spring, the bottom plate, and the funnel

I also plan to design exit chutes for each of the four coin categories

Finally, I plan to design a handle for the rotating plate so it can be easily rotated by hand (and figure out the appropriate hardware to attach it)

Bottom plate, prototype 1 vs prototype 2

Driver question 4: What mechanism will convert translational movement into the rotational movement to turn the top plate so it can distribute each coin across all 4 holes in the base plate?

Metric for Success 4: Is translational motion (back and forth or side to side) directly creating rotational motion (aka turning a gear)?

Prototype 2 Design process: I talked to others and got the idea of using a rack and pinion, and then used Leland’s link to the website where you design your own rack and pinion. I imported it into Fusion and resized it to match the size of my current top plate. Then I copy and pasted the holes in my top plate onto the new design and centered them.

I decided to redesign my spring hardstops as a slot in the rack that prevents it from moving back and forth past a certain point. I laser cut a prototype for the rack, then manually rotated it the distance it needs to move over (past all 4 holes), and measured the difference between the current slot length and the correct slot length for the hardstops. I ended up having to make the handle longer and add more teeth to the rack, because it had to turn much more than I thought it would.

Lastly, I laser cut a tiny ledge to hold the rack in place on the edge of the base plate, so the handle slides back and forth rather than moving freely.

Then I re-CADed the walls without the exit chute holes, which was much more simple.

I also removed the slots/tabs on one side of two of the walls and on one side of the base, so that there would be a smooth opening on the side with no wall. I used the dimensions tested with my friction fit template to dimension the slots and notches for my walls and base.

Then I realized I needed to suspend the base plate somehow so I could create exit chutes for the coins to drop into below it, so I decided to create a box. I changed the base plate so it would friction fit into the top of the box, and designed slots at the top of each wall for it to fit into.

4

I created my collecting bins using the CAD for my base plate to measure how far apart and how wide each section had to be, so it would align with the holes and coins would drop straight down into the correct bins (CAD pictured below, on the left)

Then I realized the coins had to fall pretty far to drop into their sections, which left room for error (sometimes they would bounce out or fall weirdly into the wrong section). I decided to add tubes that go straight down from each hole to direct the coins’ fall straight into their bins. I once again used the CAD for my base plate to create the holes, simply extruding them downwards and making them a little wider (because wider means coins wont get stuck, and I didnt want to take any chances). This worked well! (CAD pictured below, on the right)

Subsystem 3: SPRING (continued)

I measured the distance from the inner end of the spring (the hole in the top plate) to the outer end of the spring when it’s facing directly outwards from the center of the circle, and drilled a hole in my base plate at that spot. I attached the spring to both ends, and it worked!

Only later did I realize that my spring was being stretched far too much, and I had to completely rethink my design for the spring positioning.

New spring positioning!

I used the correct angles and geometry to find the right position for the hole on the bottom plate after laser cutting the top plate and measuring it all out (and marking it with sharpie). Then I drilled the hole into the bottom plate. Now the spring only moves an inch!

I also cut off half the pinion so just the necessary teeth would be included, and so it wouldn’t interfere with the screw used to fasten the outer end of the spring to the bottom plate.

Subsystem 4: FUNNEL (continued)

I reprinted the funnel to be the correct size (1 inch) and added little notches at the base with tiny holes so I could more easily attach it to the top plate (in which I made corresponding tiny holes to screw into, with heat inserts). I screwed it in, and it worked!

But then I realized I had made yet ANOTHER critical error…

REFLECTION:

Quarter stuck!

New CAD

Subsystem 2: BOX & EXIT CHUTES

Plans going forward:

Rotating plate, prototype 1 vs prototype 2

BOX WALLS:

Driver question 1: What slot length range will allow for an optimal friction fit (securely attached and removable without breaking) with my 2.1” length notches?

Metrics for Success 1: Does the 2.1” notch fit into the slot so that it stays securely but can be removed without breaking?

TUBES:

Driver question 2: What height and width does each tube need to be to get from its corresponding hole in the plate to the center of the wall from which its corresponding exit slot will be?

Metric for success 2: Does each tube get from its corresponding hole in the plate to its corresponding exit slot?

Driver question 3: What range of thickness does each tube need to be to allow coins to fall through without getting stuck?

Metric for success 3: Does each type of coin fall through its corresponding exit tube without getting stuck?

I sketched out the dimensions of each distance and used math to determine how long and wide each tube needs to be. I decided to make the thickness of the openings more than they needed to be, because there is no harm in being too thick, and I didn’t want any coins getting stuck. I measured out each tube dimension and designed them meticulously in fusion (see screenshots on the right).

I ended up scrapping this design, however, when I realized it was really hard to get a reservation at an Ultimaker, and my design was a bit too complicated for the other 3D printers. I needed to find a way to work around the tubes…

New funnel design with notches to attach to rotating plate

Test Results

The coin sorter in action!

What do I wish I had done differently?

I wish I had designed my rotation differently, so it didn’t have to rotate as far, so that I wouldn’t have to push the handle as far in. I also wish I hadn’t made my funnel incorrectly and fixed it last minute with a bracket, because that made it slightly unreliable (the bracket sometimes rotated out of place

3

1

How will my experience on this project inform decisions I will make with other designs?

You learn so much from every experience! I now know you should make templates for everything, fully think out every design before making it, and test everything a lot. Mistakes are costly. I also now fully understand the importance of creating designs that 3D printers can handle, and I’ll know what to consider for future 3D printer designs.

2

(The hole I drilled for the end of the spring)

Project brief: